16 September 20267 minute read
There is nothing wrong with a spreadsheet. It is one of the best tools ever made for thinking, modelling and getting a new process off the ground. The problem starts when the spreadsheet is no longer a working file and has become the way the business actually runs.
That shift is easy to miss. A useful tracker becomes the jobs list. A costing sheet becomes the quote. A shared workbook becomes the customer record. Before long, several versions exist, the formulas are understood by one person, and nobody is quite sure which file was sent to the client.
Where the cost hides
The licence is cheap. The operation is not. Time disappears into finding the latest version, reconciling conflicting numbers, re-keying the same information into another system, and repairing errors that only appear at month-end. Decisions are made on a copy that is almost current. New staff cannot be shown “the system” because the system is a set of habits.
- Version chaos: email attachments, desktop copies and a shared drive that never quite matches.
- Key-person risk: the workbook that only one colleague can repair.
- No reliable audit: it is hard to say who changed what, when, and why.
- Weak control: permissions are file-based, not role-based, so sensitive rows travel too far.
- Broken joins: the spreadsheet does not talk to finance, CRM or operations, so people copy data by hand.
- Ceiling on growth: what worked for one team becomes fragile when volume, sites or product lines increase.
When a spreadsheet is still the right tool
Keep using one for analysis, one-off planning, early prototypes and personal working papers. A finance lead modelling a scenario, or a manager sketching a process before it exists, should not be forced into a software project. The line is crossed when multiple people must update the same operational record as part of daily work, when customers or suppliers depend on the output, or when the file is the only history you have.
How to move on without boiling the ocean
Do not start by “building a platform to replace Excel”. Start by naming the process: who starts it, what information is required, which decisions are made, which systems already hold part of the truth, and what a finished case looks like. Then decide what should become a proper record, what can stay as a report, and what can be automated as a join between tools you already own.
Sometimes the answer is a modest operational application. Sometimes it is better use of an existing product plus integration. Sometimes it is simply agreeing a source of truth and stopping the satellite copies. The spreadsheet often remains, demoted to the job it is good at: looking at the numbers, not being the operation.
